New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Keybase errors in a relatively new team #13737

Open
RyanSinger opened this Issue Sep 13, 2018 · 15 comments

Comments

Projects
None yet
3 participants
@RyanSinger

RyanSinger commented Sep 13, 2018

Got this error: Error: ERROR CODE 218 - error unboxing chat message (transient): fast load error: team chain doesn't have a link for seqno 3, but expected one in method chat.1.local.getThreadNonblock at new r (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:467262) at file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:467791 at d (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:467830) at _rpcClient.invoke.t (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:412593) at file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2669974 at e.t.Deferrals.e._call (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2373434) at file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2373585 at r (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2373214) at e.t.Deferrals.e._fulfill (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2373544) at r (file:///opt/keybase/resources/app/desktop/dist/index.bundle.js:1:2372748)

my log id: fa1276873991301600cc651c

@RyanSinger

This comment has been minimized.

Show comment
Hide comment
@RyanSinger

RyanSinger Sep 13, 2018

Can't send messages to that team anymore

RyanSinger commented Sep 13, 2018

Can't send messages to that team anymore

@RyanSinger

This comment has been minimized.

Show comment
Hide comment
@RyanSinger

RyanSinger commented Sep 13, 2018

image

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

On it, sorry about the bug.

Contributor

maxtaco commented Sep 13, 2018

On it, sorry about the bug.

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

I think you can workaround this via keybase team list-memberships chia_network.marketing. We have a fix in the works. Sorry about this.

Contributor

maxtaco commented Sep 13, 2018

I think you can workaround this via keybase team list-memberships chia_network.marketing. We have a fix in the works. Sorry about this.

@RyanSinger

This comment has been minimized.

Show comment
Hide comment
@RyanSinger

RyanSinger Sep 13, 2018

Didn't work. See d589b9f78d729c8b0d75c41c

screenshot 2018-09-13 at 11 29 13 am

RyanSinger commented Sep 13, 2018

Didn't work. See d589b9f78d729c8b0d75c41c

screenshot 2018-09-13 at 11 29 13 am

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

Hmm, ok try a keybase db nuke and then try again?

Contributor

maxtaco commented Sep 13, 2018

Hmm, ok try a keybase db nuke and then try again?

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

Or if not that, can you click the team in the teams tab?

Contributor

maxtaco commented Sep 13, 2018

Or if not that, can you click the team in the teams tab?

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

OK, we have a fix up: #13739

We have a new "auditor" feature that randomly probes the merkle tree to make sure the server isn't trying to fork a team. It didn't work against a stale cached team. Your view of marketing is cached, and you haven't had a need to bust that cache. I think reloading team from team tab in the GUI would be enough. I thought list-memberships would, but I guess not. We had some small downtime yesterday too so you might also have missed a cache invalidation pushed from the server.

Contributor

maxtaco commented Sep 13, 2018

OK, we have a fix up: #13739

We have a new "auditor" feature that randomly probes the merkle tree to make sure the server isn't trying to fork a team. It didn't work against a stale cached team. Your view of marketing is cached, and you haven't had a need to bust that cache. I think reloading team from team tab in the GUI would be enough. I thought list-memberships would, but I guess not. We had some small downtime yesterday too so you might also have missed a cache invalidation pushed from the server.

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

@RyanSinger LMK if the workaround worked. I'd like to get a test in with the PR but we can force out a push if we have to.

Contributor

maxtaco commented Sep 13, 2018

@RyanSinger LMK if the workaround worked. I'd like to get a test in with the PR but we can force out a push if we have to.

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 13, 2018

Contributor

Should be fixed in next Linux release. Reopen if not. Thanks.

Contributor

maxtaco commented Sep 13, 2018

Should be fixed in next Linux release. Reopen if not. Thanks.

@maxtaco maxtaco closed this Sep 13, 2018

@stephthegeek

This comment has been minimized.

Show comment
Hide comment
@stephthegeek

stephthegeek Sep 14, 2018

I'm in the same team still having issues after running those commands. See screenshot: https://drive.google.com/file/d/1cW4JCVFOVO5BquZ5WfpfZa9ZPC16ZPGY/view?usp=sharing

Just submitted log id ea608df3acc33ec899538a1c

stephthegeek commented Sep 14, 2018

I'm in the same team still having issues after running those commands. See screenshot: https://drive.google.com/file/d/1cW4JCVFOVO5BquZ5WfpfZa9ZPC16ZPGY/view?usp=sharing

Just submitted log id ea608df3acc33ec899538a1c

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 14, 2018

Contributor

Gah, @stephthegeek, different bug. We are on it.

Contributor

maxtaco commented Sep 14, 2018

Gah, @stephthegeek, different bug. We are on it.

@maxtaco maxtaco reopened this Sep 14, 2018

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 14, 2018

Contributor

The issue is that a writers should be allowed to create KBFS shares, but our team player seems to be disallowing that. @stephthegeek it seems like you created the KBFS share for this team, after you were let in as a writer. We're considering 3 options: (1) relax the rules on the client; (2) kill this team; or (3) special case this link. I'm leaning toward (1), but @mlsteele and I will mull it over.

Contributor

maxtaco commented Sep 14, 2018

The issue is that a writers should be allowed to create KBFS shares, but our team player seems to be disallowing that. @stephthegeek it seems like you created the KBFS share for this team, after you were let in as a writer. We're considering 3 options: (1) relax the rules on the client; (2) kill this team; or (3) special case this link. I'm leaning toward (1), but @mlsteele and I will mull it over.

@maxtaco

This comment has been minimized.

Show comment
Hide comment
@maxtaco

maxtaco Sep 14, 2018

Contributor

Ok, fix is in master here: #13755

Question: can you guys wait 3 weeks on a release for this, or is it crucial, in which case we can (1) do a global hotfix; or (2) give you a daily (via the 'admin' channel).

Linux build should hopefully go out this afternoon

Contributor

maxtaco commented Sep 14, 2018

Ok, fix is in master here: #13755

Question: can you guys wait 3 weeks on a release for this, or is it crucial, in which case we can (1) do a global hotfix; or (2) give you a daily (via the 'admin' channel).

Linux build should hopefully go out this afternoon

@RyanSinger

This comment has been minimized.

Show comment
Hide comment
@RyanSinger

RyanSinger Sep 14, 2018

If a global hotfix is easy, go ahead. We can wait 3 weeks if needed, as the affected channel is small and we have a corresponding signal group

RyanSinger commented Sep 14, 2018

If a global hotfix is easy, go ahead. We can wait 3 weeks if needed, as the affected channel is small and we have a corresponding signal group

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment